I503
THE PHOSPHATES OF CALCIUM.
of the same concentration. Thus formaldehyde added to distilled water in the proportion of I : 40,000 had diminished to I : 160,000 in twenty-eight days, while in a parallel experiment with milk it showed the same diminution in two days. Stronger solutions are much more stable. Formaldehyde in water I : 1000 showed no appreciable loss in five months. The rate of disappearance of formaldehyde from milk containing I : 1,000 was practically the same as from water containing I : 5,000. The rate of disappearance did not seem to be affected by allowing the milk to become sour before adding the formaldehyde. The hydrochloric acid and ferric chloride test for formaldehyde in milk is sensitive to a dilution of I : 250,000. Sourness of the milk does not in itself diminish the delicacy of the reaction, but when milk is preserved by means of formaldehyde the latter will have largely disappeared before the milk becomes sour. In all of the cases here studied, however, distinct reactions were obtained a t least as long as the milk remained sweet and of normal appearance. That these results indicate a less rapid disappearance of formaldehyde than was found by Rivas or by Bigelow is thought to be due mainly to the greater delicacy of the test as modified for our experiments, though doubtless differences in laboratory temperature and in the nature of the milk used may also have affected the results. The gallic acid test is much more delicate than the hydrochloric acid and ferric chloride test and gives more conclusive results with samples which have stood until the formaldehyde has largely disappeared. QUANTITATIVELABORATORY, COLVXBIA UNIVERSITY, N E W YORK, October, 1905,
THE PHOSPHATES OF CALCIUM. BY
1.l
A. SEIDELL. Received October 16, 1905.
B. K.
CAMERON AND
WHENthe phosphates of calcium are brought into contact with water they are partially decomposed, owing to hydrolysis. At ordinary temperatures this decomposition generally proceeds a t a very slow rate, and in consequence equilibrium conditions 1
Published by permission of the Secretary of Agriculture.
1 504
F. K . CAMEROK AZJD A. SEIDELL.
have probably been attained in but very few of the experiments which have been described hitherto. The purpose of this investigation was to determine the composition of the resulting solutions and the solids in contact with them in the system water (H,O), lime (CaO), and phosphoric acid (P,O,), under conditions of final equilibrium a t a definite temperature and throughout a wide range of concentration. The temperature selected was z s 0 , and the coilcentration of phosphoric acid (P20,) varied LIP to nearly j40 granis per liter of solution, the Concentration v ith respect to lime Tar! ing in consequence to a inaximuni of 77 grams per liter. Several series oi solutions 17 ere prepared by adding so-called tricalciuni phosphate, dicalcium phosphate or monocalciurn phosphate to solutions ,f p:!osphoric acid. .lfter prolonged standing or heating, the containing bottles were rotated fur soiiie time in a nater-bath kept a t a constant teniperaturc. U71ien the contei,ts of the bottles w r e t C J be analyzed, they Trere rllon-ed to stand until the qolid niatcrinl hod settled. -1u-eighed amount oi thc solution a.; made up to definite x olume and aliquot portions were taken for analysis. P,) subtracting tllc sun1 o f ihe weights of the liiiie and phosphoric acid contained in a gil eii volume of solution from the total weight of the solution, t h c neight of the water preseiit 11 a$ cleterinineti In the analyses the lime wds determined by taking an aliquot, adding a large excess of ammonium oxalate and heating to incipient boiling Ammonia nater in excess was then added and the solution kept for some time on the hot plate a t a temperature just below boiling If the solution remained turbid, more animonium oxalate vias added until it was certain that this salt was present in excess, and a granular precipitate settled to the bottom of the vessel, leaving a clear supernatant solution. When i t appeared necessary, the filtered and washed calcium oxalate was redissolved and again precipitated in the usual way. In this way a satisfactory separation of the phosphoric acid from the lime was effected and the latter was then determined in the usual manner, weighing as oxide. The phosphoric acid was determined by the volumetric method of the Association of Official Agricultural Chemists. (
1
Bull No 46, Diviqion of Chemistry, U b Dept. Agriculture
1S99', 1)
11
THE PHOSPHATES OF CALCIUM.
1 50.5
As equilibrium is reached a t quite a slow rate, several series of mixtures were made by different methods. Series A .-To solutions of phosphoric acid of concentrations up to about 2 0 0 grams per liter, Kahlbaum’s tricalcium phosphate was added until a solid residue was left which did not dissolve even after long standing. Series B.-To solutions of phosphoric acid of the above concentration dicalcium phosphate was added until a solid residue was left. Serics C.-Monocalcium phosphate containing an excess of phosphoric acid was added to water until a solid persisted. The results of this last series have already been pub1ished.l Owing to a typographical error in misplacing a decimal point the results for calcium, as printed, are one-tenth the correct value. Making the correction, it is possible to calculate approximately, from the data given, the ratio of phosphoric acid to calcium in the solid residue, although exact ratios cannot be calculated since the specific gravities of the solutions were not determined. The identification of the solid phases in contact with the solutions was an essential feature of the investigation. For this purpose microscopic examinations were made in some instances, but the chemical analyses were made in all cases. The solids could not be separated from the adhering solution, and so portions of the solids, together with the adhering liquid, were removed, weighed and analyzed for the amounts of lime and phosphoric acid present. The water was then known by difference. The composition of the solid phases in contact with the solutions was determined by an algebraic formulation. Suppose P to represent the total phosphoric acid found in the solid residue and adhering solution, P, the phosphoric acid actually in the solid, P, the phosphoric acid dissolved in the adhering solution, and I,, I,,, I,, the corresponding values for lime. Then P = PI P, (1) and
+
+
L = L, L,.
(2)
Suppose W represent the total water found in the solid and adhering solution, W, the water combined in the solid, and W, the water in the adhering solution, Then W =W, 4- IN,. (3) 1
This Journal, 36, 1454(1904).
I 506
F. K. CANEROX AND A. SEIDELL.
If the ratio between the water and the phosphoric acid in the solid WllPl be taken as m and the ratio to lime TV1,Llbe taken as n (m and n being constants readily calculated from the formula of the substance) then substituting in ( 3 ) , it follows that or Then
and
where P,, L3 and W, represent the phosphoric acid, lime and water respectively in any given volume of the supernatant solutions, whose compositions have been determined. Substituting ( I ) and (4) in (6), it follows that
P-P, P, _ TV -mP, W, whence
p
=
PW, -P,W - - _ ~ W3-mP,
and similarly
The absolute values of PI and L,are of little interest, but their ratio will determine the nature of the solid in the residue, hence
P, (Pw,--P,w) (w,--L,) - - __ L, (W,-mP,) (LW,-L,W)' For monocalcium
phosphate [CaH,(PO,),.H,O],
n = 0.9643,
m = 0.3800 and P,/L, = 2 . 5 ; for dicalcium phosphate (CaHPO,.
aH,O), n=0.8035, m=o.6338 and P,/L,= 1.3; for tricalcium phosphate [Ca,(PO,)z], n and m disappear and P,/L, =0.8 j. These ratios are sufficiently wide apart in magnitude to make the identification of the corresponding phosphates an easy matter, and consequently in applying the above formula to the experi-
T H E PHOSPHATES OF CALCIUM.
I507
mental data only four-place logarithms were used and in the tabulated results the ratios are given only to the first decimal place. It will be seen from the table that the analyses of the solid phases in contact with the solutions yielded a constant ratio (P,O,: CaO) of about 1.3, which is the ratio for dicalcium phosphate, CaHP0,.2H,O. That equilibrium conditions had been realized was assured, for in Series A the solid initially added t o the acid solutions was tricalcium phosphate, and finally the solid was dicalcium phosphate; and in Series C the solid initially used was monocalcium phosphate, and finally the solid was dicalcium phosphate; while in Series B, dicalcium phosphate was dissolved in phosphoric acid solution and the solid in excess again proved to be dicalcium phosphate. Thus starting from solids in which the ratio of P,O,to CaO is greater than, equal to, or less than 1.3, the ratio finally obtained in all cases was 1.3, showing that a solid containing P,O, to CaO in this ratio is the stable phase in each case, and i t is safe to conclude that dicalcium phosphate is the solid phase over the range of concentration in the three series. It should be further pointed out that the points on the diagram, in which the abscissas are CaO per liter and ordinates P,O, per liter, all lie on a smooth curve, another indication that equilibrium had been reached. Two further series of solutions were prepared, using more concentrated solutions of phosphoric acid than those employed in the first three series. In one case, Series D , the solid added initially was monocalcium phosphate, while in the other case, Series E, the solid added was dicalcium phosphate. The analyses of the solid phase show that in every case it was undoubtedly monocalcium phosphate [CaH,(PO,),.H,O], the ratio of P,O, t o CaO varying but little from the calculated ratio 2 . 5 . That equilibrium had been reached was indicated by the fact that the points obtained from two differently prepared series all lay on a smooth curve; also in Series D monocalcium phosphate goes into solution as such and the solid phase remaining has the same composition as the solid originally added, while in Series E there is a complete change in the composition of the solid from dicalcium phosphate to monocalcium phosphate. The results obtained from the several series are given in the following table, and the concentrations of the solutions are represented in the accompanying chart.
I508
F. K. CAMERON AND
A. SEIDELL.
COMPOSITIONOF AQUEOUSSOLUTIONS OF LIME AND PHOSPHORIC ACID. CaO per liter solution. Grams.
Series.
A , . . . . . . . .. . . . , . , . . . .
p205 Pef liter solution. Grams.
1.71 7.48 A , . . . . . . . . . . . . . . . . . 8. ..1 0 A . . . . . . . . . . . . . . . . . .1 1. .. 5 7 c . .. . . . , . . , . . . . . . . . . . 12.88 A , . . . . . . . . . . . . . . . . 18.77 ... c . . . . . . . . . . . . . . . . . . . . 19.25 A . . . . . . . . . . . . . . . . . 23.31 ... 23.69 ..................... A , . . . . . . . . . . . . . . . . . . , 32.41 A. . . . . . . . . .. . . . . 35.90 x . .. . . . . , . . . . . . . . . . . . 39.81 . . . . . . . . . . . . . 40.89
4.69 ' 39 23.3; 36. '4 41 ' 24 59.35 63.03 75.95 79.10 109.8 129.8 139.6 142.7
B . .. . . , , . . . . . . . . . .. .
191 . o 216. j 234.6 279.7 351.9 361. I 380.3 395.1 419.7 424.6 428.0 451.7 475.3 505.8 528.9 538.3
c.. . . . . . , . . . . . . . .. . . .
154.6
55.52 59.40 70.31 ..................... 72.30 E. , . . . . , , . , . . . . . . . . . . 69.33
..................... .....................
D . . ... . . . .. . . D .. , . . . . . . . . . . . . . . . . . D. ............ D . . , . . . . . . . .. .. .. ... . . D. .. , . , . . , ... .. . . ... .. . .
.
.
.
.
.
.
.
.
.
I
.
22
53.59 48.78 44.52 41.86 39.89
Ratio
P206/C?0 in solid. 1.2
1.4 1
. ,?
1.3 I .5 1.3 1.4 1.2 1.5
1.3 1.3 1.3 1.3 1.3 1.4 1.3 1.4 1.5 2.4 2.4 2.4 2.4 2.6 2.4 2.6 2.2
2.2
2.3
2.4 2.5
When the results for concentrations in solution are plotted (see Fig.) it will be seen that they fall upon two curves. The curve representing the series of solutions in contact with dicalcium phosphate as solid phase, shows an increasing concentration of lime with increasing concentration of phosphoric acid, while on the contrary the curve representing the series of solutions in contact with monocalcium phosphate as solid phase, shows a decreasing concentration of lime with increasing concentration of phosphoric acid. Both curves approximate straight lines, although there is evident a slight curvature in the one representing the solutions in contact with dicalcium phosphate. The intersection of the two curves a t which point both mono- and dical-
THE PHOSPHATES OF CALCIUM.
Grams
CaO p e r / / ' f e r
I509
I510
F. K . CAMERON AND A. SEIDELL.
cium phosphate are stable solid phases in contact with the solution, is found by interpolation to represent a concentration of approximately 77 grams lime and 317 grams phosphoric acid per liter of solution, the solution having a specific gravity of approximately 1.29 as compared with pure water a t the same temperature. It has been noted by many investigators that when monocalcium phosphate is added in large amount to water, agitation, long standing, or rise of temperature will cause a precipitate of dicalcium phosphate to form. This fact suggests the possibility of monocalcium phosphate existing in meta-stable equilibrium in contact with a solution containing less phosphoric acid than that required by the triple point as charted. Experiments showed that such meta-stable conditions could be obtained rather readily, but no attempt was made to see how far the monocalcium curve could be extended. The tw-o dotted lines on the diagram indicate the systems in which the ratio P,O,: CaO is the same as in the compounds CaHPO, and CaH,(PO,),, respectively. The line A does not cut the curve representing the solution in equilibrium with solid dicalcium phosphate and hence it is impossible a t the temperature 25' to obtain a solution in equilibrium with a solid of the composition CaHP0,.2H20. On the other hand, the line B cuts the curve representing the solutions in equilibrium with monocalcium phosphate, but the point of intersection is in a field unstable with respect to dicalcium phosphate. It is therefore possible, theoretically, to obtain a solution of monocalcium phosphate in equilibrium with monocalcium phosphate of the composition CaH,(PO,),.H,O. Tricalcium phosphate, when brought into contact with water, always gives a solution with an acid reaction,' this fact having been observed many years ago by Warington.z Repeated washing of the solid residue finally leaves a product which yields a solution neutral to litmus, but containing a smaller ratio of P,O, 1 Apatite, of which a number of specimens from different sources have been examined, always gives a solution with a n alkaline reaction. But many specimens of South Carolina and Tennessee rock phosphate give solutions with acid reaction to litmus, a s does also the precipitated product of the laboratory. It is further worthy of note that when these phosphates which give a n acid solution a r e powdered or precipitated in a finely divided form and are shaken up with a solution of litmus, on settling they carry down all the coloring-matter with them. leaving the supernatant solution clear and colorless, though still quite acid, as can be seen by adding more litmus to the decanted liquor. 2 J. Chem. Sac., a6, 983 (1873).
THE PHOSPHATES OF CALCIUM.
1511
to CaO than that required by the formula Ca,(PO,),. It is possible to precipitate products containing still smaller ratios of P,O, to CaO which yield solutions containing more lime than phosphoric acid and which are alkaline to such indicators as litmus or phenolphthalein. Some preliminary attempts were made to follow the solubility curve for lime in solutions of phosphoric acid of lower concentrations than are given in the table, but owing to the length of time required to reach conditions of equilibrium and the analytical difficulties involved, the reliability of the results obtained was open to doubt and consequently this work was postponed for a separate investigation. Some idea as to the concentrations below which tricalcium phosphate might exist as a stable solid phase was, however, obtained. A rather dilute solution of sodium phosphate, alkaline to litmus or phenolphthalein, was added to a dilute neutral solution of calcium chloride. After the addition of the first drop or two of the alkaline phosphate solution, the neutral calcium chloride solution became acid, and became more and more acid on further addition of the alkaline solution, as indicated by the color of the added litmus, up to a certain point, then gradually less and less acid, finally becoming alkaline. The points a t which the maximum acidity was reached and a t which the solution passed from acid to alkaline could not be sharply determined. The precipitated solid obtained by adding various amounts of the phosphate solution to a definite volume of the calcium chloride solution was always found to be dicalcium phosphate whenever the resulting supernatant solution possessed an acid reaction. When the supernatant solution was alkaline, however, the precipitate always contained a larger proportion of lime than is required by the formula for tricalcium phosphate. It became evident, therefore, that either tricalcium phosphate can exist under equilibrium conditions in contact with water over but a very narrow range of concentration or that it exists only as a metastable form in contact with water a t room temperature, although the rate of change t o a more stable form is probably quite slow. The results given in the foregoing paragraphs have a special significance for agricultural investigations. It is generally held that monocalcium phosphate is the form contained in superphosphates or “acidulated” phosphates, and that the dicalcium phosphate is the salt of “reverted” phosphates. It is evident,
1512
F. K. CAMERON AND J. M. BELL.
however, that no such concentration of the soil moisture with respect to phosphoric acid can be attained as is necessary for the continued existence of monocalcium phosphate, and that i t must promptly and quite completely ‘revert” nhen added to the soil. It is obvious, therefore, that the relative values of the two phosphates as manures can not be entirely accounted for by the marked difference in their respectix e solubilities and consequent “availability,” as has often been supposed. BLKhAL O x
L
Z
5011s
DE I !KTVI \T O F A G R I C L I TI \\A\AI\G1O\ D C
RE
THE PHOSPHATES OF CALCIUM. 11.’ 13s ’r
I; C I M L R O N A Y D J
SI. BE1 I .
Received October 16, rgaj.
IN THE foregoing paper the equilibrium conditions in tli e system liine (Ca0)-phosphoric acid (P,O,)-n ater, have been found for solutions which are comparatix-el! quite concentrated with respect to both lime and phosphoric wid. It has been shown that for concentrations of P,0, abox-e 21; grains per liter the stable solid phase a t 25‘ C is nionocalciuni phosphate, and below that concentration down to 20 grams per liter the solid phase is dicalcium phosphate. The object of the present irvestigatiori as to find the equilibrium conditions for still lon cr concentrations of phosphoric acid and of l i m and. especiallJ- in vien- 01 some preliminary experiments hy Llr. Seidell. to find whetlier there might be a series of solutions, all of uhich can exist in equilibrium with tricalcium phosphate Several series of solution< n ere prepared b! adding lime-w ater (Series 11) or dicalciurii phosphate (Series S) to solutions of phosphoric acid. The containing bottles were rotated in a constant temperature bath for about three months. After the analyses of the solutions had been made, i t was found that between the end solutions of the foregoing paper and the most concentrated solutions of the present series there was a very large range for which there were no data. Another series (Series P) was prepared by adding dicalcium phosphate to somewhat more concentrated solutions of phosphoric acid than had been used in Series N. These solutions were rotated in the constant temperature bath 1
Published by permissioii of the Secretary of Agrlculture.
